The start to 2025 has, yes, been volatile.
In fast-moving times like this, predictability holds a special kind of appeal. In my field of software monetization, I see predictability taking on a new kind of importance.
Predictability, adaptability, and resilience are always important to running a business, but can become more challenging during an era of volatility. Software producers and suppliers need to market and sell products that meet the needs of their customers, while also supporting their businesses’ revenue goals. Software monetization needs to play a facilitating role in this equation. Customers want clarity into how much they’re going to spend on software. Suppliers want to be able to accurately forecast revenue and adjust their offerings to shifts in the market.
When sales and marketing teams wonder, “How do we make more money off what our business is doing?,” part of the answer is to ensure that software packaging decisions and monetization approaches align with corporate–and customer–needs. The packaging choices and data supporting the operational aspects of the sale and customer relationship must offer sufficient granularity and visibility into spend and value.
Today, multiple widely-used monetization models offer the ability to meet these needs with various levels of preference and suitability. As the market shifts, it’s crucial to understand how they function and why combining approaches to offer a hybrid approach to monetization can help deliver more predictable software revenue.
Monetization Models for Predictable Income
Preferences for how to sell and monetize software have evolved over recent years. Perpetual licenses—where a user pays once for unlimited use of a software product—have long been popular and still remain in widespread use. But there have been a few noteworthy shifts in software monetization in the past twenty years that have emerged from a combination of a shift for primarily hardware-based revenue to software and an increase in wanting to adjust the provider-customer relationship from a one-time event to an ongoing relationship.
Subscription monetization, in which a user pays monthly or yearly to use software, increases the connectivity between the producer and its customers. The cost of entry is often appealing for a user who wants access to the software for a brief period of time. When a user maintains a subscription over time and continuously renews, the profitability of the product increases, offering predictable income and annual recurring revenue (ARR). The customer also has clarity into exactly what their software expense will be, whether using it for a single month for a niche project or maintaining an ongoing subscription to support core business needs. Subscription offerings have surged, thanks in part to how well suited they are for software-as-a-service (SaaS) deployments. Today, 29% of software suppliers use subscription/term monetization extensively, with another 59% using it moderately, as reported in the Revenera Monetization Monitor: Software Monetization Models and Strategies 2025 Outlook.
A relatively new player in the software monetization field is usage-based monetization models–defined by a usage metric, such as consumption, credit-based tokens, elastic access, metered, or pay-per-use. The increasing popularity of usage-based approaches comes from their ability to help software suppliers align per-use costs with revenue, such as for artificial intelligence (AI) offerings.
Customers want to know how much they’re spending and have clear insights into what they’re using. The specific implementation of a usage-based model can facilitate this. Consider a monetization model based around pre-purchased tokens. A token is a flexible entitlement to a set of technology offerings. Each offering within a software producer’s portfolio of products has an associated value token that is managed through a rate table. Customers can budget around how much to spend on the number of tokens they estimate will meet their needs, establishing a predictable expense, then use those tokens when needed.
When combined with other purchased licenses under a perpetual or subscription plan this can be useful for preventing spikes in expenses in situations such as during peak times, or for particular projects when a team member who doesn’t have a subscription or perpetual license needs access to a particular product (or feature within it). Rather than having to purchase an entitlement for more than that individual needs, that person can use tokens to access the product(s)/feature(s) required in the moment. Tokens help suppliers manage costs. Expensive features (such as for IoT devices and AI compute) are often more costly than other standard features; this can be reflected in the token rate table.
Monetization Approaches for Emerging Technologies
Predictable revenue from AI features and the software-driven functionality of IoT devices relies on accurate measurements of the expenses associated with these emerging technologies.
Suppliers must have clarity into:
- How much is being spent to produce the software?
- How can a software producer accurately predict expenses (including cloud costs), so as to not underprice software products, whether via subscription or consumption-based models?
- What’s the actual expense associated with that usage and how can a software supplier offset some of the costs?
- What happens if a customer is regularly using high-cost features?
In recent years, companies have been trying to address the true expenses associated with AI. For example, a search performed with a large language model (LLM) costs about 10 times more than a standard keyword search. If even Google is struggling to clearly measure and reflect the impact of this spiked expense, what does it mean for all other companies that grapple with the costs associated with AI?
Those expenses spike across AI functionalities, such as creating text, images, or music. Why? Queries hit processors hard for a short period of time. When that hit happens, so does the expense to the software producer. If a subscription customer relies heavily on AI, the supplier’s profit will drop for that user for that term—impacting the bottom line and the overall ability to predict margin.
Generally speaking, software providers don’t want to carry that extra expense. They need to charge it through, which is why consumption models are the standard for AI tools. Usage-based models offer an important approach to cost management for AI providers and for the IoT device vendors who are looking to generate revenue from the rapidly growing number of intelligent devices.
With the consumption model, software suppliers have the ability to track when usage events happen and the duration of that use. Both of these dimensions are needed to accurately measure and bill for usage. The true costs can be reflected in the rate table, where a token for AI-driven functionality may be multiple times the cost of a more traditional piece of functionality within the product.
Plan for Predictability
For you and for your customers, predictability plays an important role in business decisions. Hybrid approaches to monetization offer great flexibility to support your goals. Be clear on what you need and what best meets each customer’s needs. Your bottom line—and your customer relationships—will benefit.
About the Author
Eric Jensen, a solution architect whose focus is on software monetization transitions, has been with Revenera for nearly 25 years.
